Neonatal Seizure detection using GLCM feature extraction & AlexNet classification

• Published in: Multimedia tools and applications Vol. 83; no. 35; pp. 83139 - 83155
• Main Authors: Jebin, Ben M., Rejula, M. Anline, Eberlein, G.
• Format: Journal Article
• Language: English
• Published: New York Springer US 01.10.2024; Springer Nature B.V
• Subjects: Algorithms; Brain; Classification; Computer Communication Networks; Computer Science; Convulsions & seizures; Data Structures and Information Theory; Decision trees; Electroencephalography; Error detection; Feature extraction; Fourier series; Multimedia Information Systems; Neonatal care; Neurological diseases; Seizures; Signal classification; Special Purpose and Application-Based Systems; Statistical analysis; Statistical methods; Track 2: Medical Applications of Multimedia; Electroencephalography (EEG); Gray-level co-occurrence matrix (GLCM); Recurrent Neural Network (RNN); Convolutional Neural Network (CNN)
• ISSN: 1573-7721; 1380-7501; 1573-7721
• DOI: 10.1007/s11042-024-18779-8

Neonatal seizures result from a decreased supply of oxygen to the brain during the neonatal period, leading to neurological disorders and abnormal neuronal activity. Electroencephalography (EEG) is employed to monitor brain signals, and Fourier series and transformations are applied to analyze the frequency of these signals. Feature extraction is performed using the grey-level co-occurrence matrix (GLCM), a statistical method that analyses EEG images by scaling them between 0 and 1 before applying a classifier algorithm. Detecting neonatal seizures is particularly challenging due to the complexity of understanding EEG signals in neonates. Various classification algorithms, such as K-Nearest Neighbour (K-NN), Naïve Bayes, Logistic Regression, Decision Tree (D.T.), and Random Tree, are employed to identify seizure occurrences from EEG signals. The proposed classification approach leverages the AlexNet algorithm, demonstrating superior performance and accuracy compared to existing classification methods in the detection of neonatal seizures. The efficiency of the proposed methodology lies in its ability to train the machine rapidly, achieving a reduced training error rate of 25%, facilitated by the incorporation of maxpooling and softmax layers in the classification algorithms. The overall accuracy of the system reaches 94%, with a specificity of 96%, enabling automated and effective detection of neonatal seizures.